A ‘tsunami’ on a silicon chip

Scientists have created a tiny tsunami of light that could be used on a silicon chip. But rather than the destruction usually associated with a tsunami, they plan to use this one to make smaller, faster signals and more energy efficient processors.

Tsunamis are an example of a soliton, a pulse of wave energy that does not spread out, maintaining its shape as it travels. 

As well as their ability to travel long distances unchanged, solitons can be used for complex signal processing, such as generating trains of pulses, pulse shaping and compression.

“It opens up an entire new toolbox of pulse manipulation techniques,” says Ben Eggleton from Australia’s University of Sydney. “We rely on pulses of light for all kinds of communications technology.” 

The report of Eggleton and his team’s collaboration with Singapore University of Technology and Design is published in the journal Laser and Photonics Reviews. 

Although optic fibres are used to carry signals huge distances, much of the signal processing involved with transmitting and receiving data is still done with electronics. 

Eggleton’s dream is to not have to convert light to electricity, but instead to find ways of processing light signals, a field known as photonics.

Signal processing using soliton dynamics has been possible, but only by sending the signal through a long length of optic fibre, negating the speed advantages of light processing. 

To create a soliton, one needs the refractive index – a measure of the speed of the wave – to vary in such a way that the centre of pulse, where the intensity is highest, travels slightly more slowly than the edges of the pulse. This allows the edges to keep up, so that the pulse does not spread out. 

This requires a fine balance between two properties: the intensity dependence of the speed (nonlinearity) and frequency dependence of the speed (dispersion).

Few materials have the right combination, but high dispersion can be created in an optic fibre by surrounding it with a regular structure known as a Bragg grating.

However the high intensity required to reach the nonlinear regime makes things tricky: for example, silicon, the most common material for photonics, absorbs a lot of energy when the signals reach the intensity needed to create a soliton, ruling out its use.

When Eggleton visited Singapore a few years ago, he saw Ezgi Sahin creating “exquisite” Bragg gratings with lithography: not with silicon, but silicon nitride (Si3N4), a material that does not have the energy loss problem of silicon.

“It was a real Eureka moment,” Eggleton says.

The two institutions went on to collaborate and demonstrate solitons in silicon nitride, finding the best material is a silicon-rich silicon nitride.

Eggleton says a soliton-based signal processor can now be developed that can sit on a chip and be integrated with other photonic components such as lasers and detectors.

“It’s the holy grail, to put everything onto a chip,” he says.